Comparing nonparametric k-nearest neighbor technique with ANN model for predicting soil saturated hydraulic conductivity

Soil saturated hydraulic conductivity is the most important physical parameter, but its measurement often is difficult because of practical and/or cost-related reasons. In this research, expert system approaches with one type of the nonparametric lazy learning algorithms, a k-nearest neighbor (k-NN) algorithm, was compared and tested to estimate saturated hydraulic conductivity (Ks) from other easily available soil properties. In this research 151 soil samples were collected from farms land around Bojnourd and saturated hydraulic conductivity (Ks) was estimated from other soil properties including soil textural fractions, EC, pH, SP, OC, TNV, ρs and ρb. Results showed that the accuracy of the parameter estimation, using parametric method of artificial neural network to compare with k-nearest neighbors for terms of all the parameters (with r=0.97, EF=0.946, RMSE= 8.798, ME= 28.446 and CRM =-0.144) compared to other methods input models is acceptable and can used to estimate saturated hydraulic conductivity especially when for new data set available these functions is essential.